Since drugs are usually available to humans either simultaneously or sequentially with other reinforcers, an understanding of factors that control choice between drugs and other drug or non-drug reinforcers is important to a complete understanding drug abuse. Those factors might be environmental, pharmacological or organismal in nature. The experiments described in the present proposal will utilize an animal model of drug abuse, i.v. self-administration, to accomplish two major goals: l) to continue our ongoing investigation of environmental and pharmacological determinants of. drug choice; and 2) to examine the applicability to drug choice of models of choice that have evolved from the experimental analysis of behavior: behavioral economics, maximizing and matching. Although all three models have been shown to account for substantial amounts of data from experiments utilizing non-drug reinforcers, there has been little or no research designed to establish the generality of these models to drugs as reinforcers. The research has three Specific Aims that are addressed in three series of experiments. Specific Aim 1 is to extend our ongoing investigation of response requirement as a determinant of drug choice using FR schedules of reinforcement and to test the unit price component of the behavioral economics model in the context of choice. Specific Aim 2 is to examine maximization as an account of drug choice using discrete-trials probability choice schedules of reinforcement. Specific Aim 3 is to utilize concurrent VI schedules of reinforcement to assess the applicability of the matching law to drug choice. In each series of experiments, we will initially study choice between two injections of the same drug to establish determinants of choice when the nature of the alternative reinforcer is not an issue. To examine the pharmacological generality of our results, we will study a representative of each of the three major classes of abused drugs: cocaine (psychomotor stimulant), methohexital (barbiturate) and alfentanyl (opioid) across a range of doses. Since drug self-administration is one of many options available to the drug user, it is important to establish the generality of our findings across a range of alternative reinforcers. Therefore, we will next examine in all three models the role of the nature of the alternative reinforcer in cocaine choice. We will continue our investigation of choice between cocaine and food, a biologically necessary non-drug reinforcer, and extend that analysis to include choice between cocaine and other drug and non-drug reinforcers. A complete understanding of the determinants of drug choice would suggest approaches to modifying that choice. Moreover, establishing the applicability of the different models of choice to drug self- administration would extend beyond the single operant situation the generality of the conclusion that drugs maintain behavior in the same manner as non-drug reinforcers.